Re: [lmi] MinGW-w64 gcc-7.2 -Wconversion warning in wx

[Vadim Zeitlin]

On Sat, 3 Mar 2018 01:38:44 +0000 Greg Chicares <address@hidden> wrote:

GC> I'll turn this warning off for now with '-Wno-conversion', because
GC> it's triggered here in the version of wx we're currently using:
GC> 
GC> /opt/lmi/local/include/wx-3.1/wx/gdicmn.h:322:21: error: conversion to 
'float' from 'int' may alter its value [-Werror=conversion]
GC>          { x = (int)(x*xscale); y = (int)(y*yscale); return *this; }
GC>                      ^
GC> /opt/lmi/local/include/wx-3.1/wx/gdicmn.h:322:42: error: conversion to 
'float' from 'int' may alter its value [-Werror=conversion]
GC>          { x = (int)(x*xscale); y = (int)(y*yscale); return *this; }
GC>                                           ^
GC> 
GC> It would be nice to fix that (if it hasn't already been fixed in
GC> a more recent version) because this warning is triggered by some
GC> code in lmi (which I haven't investigated yet).

 This warning will be fixed once https://github.com/wxWidgets/wxWidgets/pull/754
is merged (which I'll do soon), but there are many other -Wconversion
warnings in lmi code itself, at least when building with autotools-based
system. First, there were many, many -Wfloat-conversion warnings, e.g.

ihs_acctval.cpp:1122:45: error: conversion to ‘double’ from ‘long double’ may 
alter its value [-Werror=float-conversion]
         double const z = material_difference(MinInitPrem(), 
InitAnnPlannedPrem_);
                          
~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

so I had to add -Wno-float-conversion to the compilation flags. After this
only the following warnings remain:


facets.cpp: In instantiation of ‘static const mask* 
{anonymous}::C_is_not_whitespace_ctype<C>::get_table() [with unsigned char C = 
32; std::ctype_base::mask = short unsigned int]’:
facets.cpp:74:40:   required from 
‘{anonymous}::C_is_not_whitespace_ctype<C>::C_is_not_whitespace_ctype() [with 
unsigned char C = 32]’
facets.cpp:90:22:   required from ‘const std::locale& 
{anonymous}::locale_with_facet() [with Facet = 
{anonymous}::C_is_not_whitespace_ctype<32>]’
facets.cpp:100:62:   required from here
facets.cpp:82:19: error: conversion to ‘std::ctype_base::mask {aka short 
unsigned int}’ from ‘int’ may alter its value [-Werror=conversion]
             rc[C] &= ~std::ctype_base::space;
round_glibc.cpp: In function ‘long double roundl(long double)’:
round_glibc.cpp:389:25: error: conversion to ‘short int’ from ‘int’ may alter 
its value [-Werror=conversion]
round_glibc.cpp:222:30:
   iw_u.parts.sign_exponent = (exp);                             \
                              ~~~~~
round_glibc.cpp:389:25:
   SET_LDOUBLE_WORDS (x, se, i0, i1);
round_glibc.cpp:222:31: note: in definition of macro ‘SET_LDOUBLE_WORDS’
   iw_u.parts.sign_exponent = (exp);                             \
                               ^~~
wx_utility.cpp: In function ‘wxDateTime ConvertDateToWx(const calendar_date&)’:
wx_utility.cpp:123:22: error: conversion to ‘wxDateTime::wxDateTime_t {aka 
short unsigned int}’ from ‘int’ may alter its value [-Werror=conversion]
         (lmi_date.day()
          ~~~~~~~~~~~~^~
In file included from main_cli.cpp:46:0:
timer.hpp: In instantiation of ‘AliquotTimer<F>& AliquotTimer<F>::operator()() 
[with F = std::_Bind<illustrator(const char*, Input)>]’:
timer.hpp:278:43:   required from ‘AliquotTimer<F> TimeAnAliquot(F, double) 
[with F = std::_Bind<illustrator(const char*, Input)>]’
main_cli.cpp:139:63:   required from here
timer.hpp:216:14: error: conversion to ‘double’ from ‘elapsed_t {aka long long 
unsigned int}’ may alter its value [-Werror=conversion]
         ;now < expiry_min || j < 100 && now < expiry_max
          ~~~~^~~~~~~~~~~~
timer.hpp:216:45: error: conversion to ‘double’ from ‘elapsed_t {aka long long 
unsigned int}’ may alter its value [-Werror=conversion]
         ;now < expiry_min || j < 100 && now < expiry_max
                                         ~~~~^~~~~~~~~~~~
timer.hpp:229:26: error: conversion to ‘double’ from ‘elapsed_t {aka long long 
unsigned int}’ may alter its value [-Werror=conversion]
     unit_time_ = minimum / dbl_freq;
                  ~~~~~~~~^~~~~~~~~~
rate_table.cpp: In member function ‘void 
binary_format::writer::write({anonymous}::enum_soa_field, const 
std::experimental::fundamentals_v1::optional<std::__cxx11::basic_string<char> 
>&)’:
rate_table.cpp:528:69: error: conversion to ‘uint16_t {aka short unsigned int}’ 
from ‘std::__cxx11::basic_string<char>::size_type {aka unsigned int}’ may alter 
its value [-Werror=conversion]
         do_write_record_header(soa_fields[field].record_type, length);
                                                                     ^
rate_table.cpp: In member function ‘void 
text_format::writer::write_values(const std::vector<double>&, const 
std::experimental::fundamentals_v1::optional<short unsigned int>&, const 
std::experimental::fundamentals_v1::optional<short unsigned int>&, const 
std::experimental::fundamentals_v1::optional<short unsigned int>&, const 
std::experimental::fundamentals_v1::optional<short unsigned int>&, const 
std::experimental::fundamentals_v1::optional<short unsigned int>&)’:
rate_table.cpp:655:58: error: conversion to ‘uint16_t {aka short unsigned int}’ 
from ‘int’ may alter its value [-Werror=conversion]
         for(std::uint16_t age = *max_select_age + period + 1; age <= *max_age; 
++age)
                                 ~~~~~~~~~~~~~~~~~~~~~~~~~^~~
rate_table.cpp: In member function ‘double 
soa_v3_format::table_impl::parse_single_value(const char*, const char*&, int)’:
rate_table.cpp:1560:29: error: conversion to ‘double’ from ‘long long unsigned 
int’ may alter its value [-Werror=conversion]
         return res_int_part.num;
                ~~~~~~~~~~~~~^~~
rate_table.cpp: In member function ‘void 
soa_v3_format::table_impl::parse_values(std::istream&, int&)’:
rate_table.cpp:1776:25: error: conversion to ‘short unsigned int’ from ‘int’ 
may alter its value [-Werror=conversion]
                 age += *select_period_;
                         ^~~~~~~~~~~~~~
rate_table.cpp: In member function ‘void soa_v3_format::table_impl::validate()’:
rate_table.cpp:1855:72: error: conversion to ‘uint16_t {aka short unsigned 
int}’ from ‘std::size_t {aka unsigned int}’ may alter its value 
[-Werror=conversion]
         std::uint16_t required_num_decimals = 
deduce_number_of_decimals(values_);
                                               
~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~
rate_table.cpp: In instantiation of ‘T {anonymous}::get_value_or(const 
std::experimental::fundamentals_v1::optional<_Tp>&, U) [with T = short unsigned 
int; U = int]’:
rate_table.cpp:618:37:   required from here
rate_table.cpp:178:14: error: conversion to ‘short unsigned int’ from ‘int’ may 
alter its value [-Werror=conversion]
     return o ? *o : v;
            ~~^~~~~~~~


 None of them look really difficult to fix, but I'm not sure if it's worth
doing this, what do you think?

VZ